基于贝叶斯分类的变压器绕组故障诊断算法  

Fault Diagnosis Algorithm of Transformer Windings Based on Bayesian Classification

作　　者：李学生[1] 张尊扬[1] LI Xuesheng;ZHANG Zunyang(School of Electrical and Information Engineering,North Minzu University,Yinchuan 750021,Ningxia,China)

机构地区：[1]北方民族大学电气信息工程学院,宁夏银川750021

出　　处：《济南大学学报（自然科学版）》2021年第4期412-416,共5页Journal of University of Jinan(Science and Technology)

基　　金：国家自然科学基金项目(51867001);宁夏回族自治区重点研发计划项目(2019BDE03010)。

摘　　要：为了提高变压器绕组故障诊断准确率,减少不同故障类型的诊断时间,提出基于贝叶斯分类的变压器绕组故障诊断模型;根据平行板电容理论,计算电容器极板上所带电量,利用电场分布和电场能量,计算导体间的互电容之和,将绕组电感和电阻确定为贝叶斯诊断模型的故障特征量;利用后验概率决策分类,通过欧拉非线性理论模型离散获得五阶故障模型,构建故障模型的输出方程,完成基于贝叶斯分类的变压器绕组故障诊断。结果表明,该方法的故障诊断准确率可高达96%,针对不同的故障类型诊断的实时性较好。To improve accuracy of transformer winding fault diagnosis and reduce diagnosis time of different fault types,a transformer winding fault diagnosis model based on Bayesian classification was proposed.According to parallel plate capacitance theory,electric quantities on capacitor plates were calculated.Electric field distribution and electric field energy were used to calculate the sum of mutual capacitance between conductors.Winding inductance and resistance were determined as the fault characteristic quantities of Bayesian diagnosis model.A fifth-order fault model was obtained by using posterior probability decision classification and discretization of nonlinear Eulerian theory.An output equation of the fault model was constructed,thereby the transformer winding fault diagnosis was completed based on Bayesian classification.The results show that the fault diagnosis accuracy of the proposed method can be as high as 96%,and the real-time performance of fault diagnosis is good according to different fault types.

关 键 词：贝叶斯分类 绕组电感 绕组电阻 故障特征量 

分 类 号：TM418[电气工程—电器]